Reproduce “Using Phonological-Level Wav2Vec2 for Mandarin Automatic Mispronunciation Detection and Diagnosis” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jinghao Chen, Mostafa Shahin, Beena Ahmed
The authors present a Mandarin mispronunciation detector that predicts sound attributes, including segmental features and tone, rather than only phoneme labels. It uses those attributes to generate feature- and phoneme-level diagnoses. On LATIC, IPA-D × Tone-CAT reports FAR of 8.15% and DER of 27.86%, versus 9.97% and 34.03% for a matched phoneme-only baseline; the evaluation has four speakers.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.